The Forest from the Trees: A post for the stats nerds


#381

I apologise to St Francis of CD for under-rating him. Still, can do better. Who’s the oldest player in the league?


#382

PAPLEY has been updated- happy trade evaluating


#383

Interesting BBB.
This shows McCluggage as having a higher future PAV than Zac Merret and Pidge ( who doesn’t even appear in the top 24).


#384

Mccluggage is four(?) Years younger than zerrett? Thats 88 fewer games left in Zerretts career. The list heavily favours younger players in general.

Their PAV system hasn’t been a fan of McGrath based as it is around productivity, hence why it LOVES Oliver who is a busy bee indeed. They acknowledge that shutdown defenders are difficult to rate. I think we are all hoping McGrath finds the pill a bit more consistently next season, lord knows we’ll be a better team for it.


#385

Interesting that SPP is higher still, yet everyone on Blitz calls him a spud.


#386

Under their system, SPP had the best first year since Dusty. Interestingly, seems to have the personality to match.


#387

Interesting that Ed Richards is so high. Started as a lock down small defender before going forward. Think any Dogs fan would rate Aaron Naughton much higher yet they have it around the other way.


#388

Last year I compared the Crichton Medal results to HPN’s PAV ranking tool - the results were surprisingly bang on. Let’s do the same for 2018:

2018 Crichton Medal top 10:

  1. Devon Smith 370 votes
  2. Dyson Heppell 360 votes
  3. Zach Merrett 335 votes
  4. Adam Saad 315 votes
  5. Anthony McDonald-Tipungwuti 295 votes
  6. Tom Bellchambers 281 votes
  7. Michael Hurley 280 votes
  8. Cale Hooker 276 votes
  9. Conor McKenna 268 votes
  10. Brendon Goddard 267 votes

HPN PAV Top 10

  1. Dyson Heppell, 19.25
  2. Devon Smith, 18.55
  3. Zach Merrett, 18.27
  4. Tom Bellchambers, 17.53
  5. Jake Stringer, 16.01
  6. Athony McDonald-Tipungwuti, 13.58
  7. Michael Hurley, 13.43
  8. Cale Hooker, 13.1
  9. David Myers, 12.87
  10. David Zaharakis, 12.66
  11. Adam Saad, 12.09

#389

Some observations:

  • We are missing Daniher who was ranked 12th in the competition in 2017.
  • Heppell and Zerrett were marginally down on their 2017 output.
  • Goddard’s drop off was significant, from an almost elite PAV ~15 in 2017, to just over 10 in 2018. This without a significant role change.
  • Smith’s year was his best by a mile, his closest being 2015 with a near 15 PAV
  • Stringer had a brilliant year, almost as valuable as his 2015 AA year. About a third of his value came from his midfield work. A few more goals and centre clearances will see him in genuinely elite company.

#390

One more, Myers had a decent year but down slightly on his 2014 output.


#391

It’s a pretty thorough system they have developed. Certainly shows some good guidelines for trading players.

Interesting that over time picks in the top 19 are considered to have a distinctive higher long term value but after that it a far different value change.

I guess this would be an excellent measure of the relative worth of a draft if the quality runs deeper past 19 or before.

I’ll have to go and check if he has done a correlation of pre draft PAPLEY vs post draft using TAC cup data?


#392

They haven’t done anything using TAC data that I’m aware of.

In footballistics they analyse each draft to highlight their strength one, two, three years in etc. Still relatively early days, but Zerretts draft is looking monumentally stacked. Good thing the dogs wanted Crameri!

The other takeaway from their system is the value of late 2nd round picks. There really isn’t a huge difference in historical value between say pick 22 and pick 39. Greasing the trade wheels, like we did, with a downgrade of a 2nd rounder is probably a clever strategy


#393

How is this pav thing actually calculated


#394

METHOD

  • Step 1

Preheat oven to 200°C. Draw a 24cm (diameter) circle on a sheet of baking paper. Place, pencil-side down, on a baking tray. Dust lightly with 1 teaspoon cornflour.

  • Step 2

Using an electric mixer, beat eggwhites and cream of tartar in a large bowl until soft peaks form. Add sugar, 1 tablespoon at a time, beating constantly until thick and glossy. Add remaining 3 teaspoons cornflour with the last tablespoon of sugar. Fold through vanilla and vinegar.

  • Step 3

Spoon meringue onto baking paper. Shape into a circle, using the pencil mark as a guide, with a slightly higher edge and a low centre. Reduce oven to 100°C. Bake for 1 1/4 to 1 1/2 hours or until dry and crisp. Turn off oven and open oven door. Cool completely in oven (pavlova may sink during cooling).

  • Step 4

Slide pavlova onto a serving plate. Spread with cream and top with raspberries. Serve.


#395

Capture


#396

Ohhhh SNAP!!

And also, … fk you Dunny!! :stuck_out_tongue_closed_eyes:


#397

It is broken down into forward, midfield and defense and basically tries to allocate each players portion of the overall teams outout in those areas, so if player A has a midfield pav of 10, that’s 10% of that teams total midfield Pav. How they calculate those are on the website, but it’s just using freely available stats.


#398

Ratios from HPM, showing the relative effectiveness of our top mids on a small set of ratios, with Shiel added in, compared with Mitchell.

It appears that although Shiel can play contested footy, he has been playing more of an outside, forward-mid role in 2018.

As it is our midfielder ratings would be 1. Smith, 2 Heppel 3 Shiel, 4 Merret

image


#399

Is there any chance that Langfords stats are available for those categories?


#400

So what that chart tells me is not going to be taking a lot of contested marks, or marks in general. But is the man who you want to have with the ball in hand delivering inside 50. An attacking weapon.

Given that he has also previously played more of an inside role than he did this year, then that is a good thing for his all round midfielder status.

Seems a real nice fit for us.